46. Nibble metasomatic K-feldspathization
RONG Jiashu
Email: rongjs123@vip.sina.com
Beijing Research Institute of Uranium Geology
P.O. Box 9818, 100029, Beijing, China
March 4, 2003
Revised, May 18, 2004
Nibble metasomatic K-feldspathization intensively occurred in undeformed Huangnitian hornblende quartz syenite. The main replaced mineral is the widely distributed primary K-feldspar. Nibble replacing K-feldspar occurs on the grain boundaries between two differently oriented K-feldspar crystals, takes the same crystallographic orientation as that of the primary K-feldspar on which it nucleates, and nibbles at the opposite, differently oriented K-feldspar. The nibble metasomatic K-feldspar grows as much as 3 mm while enclosing and preserving relics of perthitic albite of the replaced K-feldspar. These albite relics keep the same original orientation. Quartz and hornblende show little to no replacement during the nibble K-feldspathization. Because of the nibble metasomatic mechanism and the consistent presence of unreplaced remnants of plagioclase in the nibble replacing K-feldspar, most parts of the K-feldspar megacrysts, lacking these features, are likely primary instead of metasomatic.
The Cretaceous Huangnitian hornblende quartz syenite is located east of Xingzhou in Yangjiang County, Guangdong province, southern China (Mo Zhusun et al., 1980). It occurs at the junction of three successive intrusions into the surrounding rocks (Fig. 1). The hornblende quartz syenite, occupying more than 1 km2, is medium grained, consists of potash feldspar (60%), plagioclase (20%), quartz (8%), hornblende (10%), and minor opaque minerals. Plagioclase (An7-13) is moderately zoned, elongated, and more or less euhedral. Hornblende occurs as short prismatic crystals. Quartz is interstitial.
The appearance of K-feldspar in the rock is quite complicated at first glance under cross-polarized light, because the K-feldspar is separated into several amorphous parts with different orientation (Fig. 2a; Fig. 2b). Vermicular perthitic albite stringers and several isolated albite blebs occur in some of the K-feldspar parts (Fig. 3a). When a quartz plate is inserted (Fig. 3b; Fig.3c), each of the isolated albitic blebs (Ab1 in K-feldspar K2') can be seen to have the same crystallographic orientation as that of the perthitic lamellae (Ab1)of the adjacent K-feldspar. Therefore, the isolated albite blebs are likely remnants of perthitic lamellae of the replaced previously-formed K-feldspar. The vermicular perthitic albite stringers, having the same orientation as the host K-feldspar, are generally pointing at the border with adjacent minerals (Fig. 3b, Fig. 4b, and Fig. 5). This correlation strongly suggests that the K-feldspar (K2') is metasomatic in origin; i.e., the primary K-feldspar (K1) has been replaced by nibble metasomatic K-feldspar (K2') with different orientation.
The pre-existing K-feldspar crystals are probably magmatic and primary in origin because they are subhedral ((K2 in Fig. 4d; K1 in Fig. 5) and are characterized by the presence of indistinct normal perthitic albite, the more or less homogeneous argillation, and the absence of metasomatic remnants of perthitic albite, whereas the nibble metasomatic K-feldspar is distinguished by the appearance of scattered remnants of perthitic albite of the replaced adjacent K-feldspar, the vermicular or stringer perthitic albite, and less or inhomogeneous argillation. The orientation of metasomatic K-feldspar (Kx') is different from that of the K-feldspar being replaced, but strictly coincides with the adjacent K-feldspar (Kx) on which it epitaxially grew. Because the replacing K-feldspar may extend as much as 3 mm into one primary K-feldspar being replaced in front, another primary K-feldspar on which the replacing K-feldspar nucleated naturally may not appear together in one thin section.
Swapped nibble metasomatic K-feldspar (K1', K2'), however, can still fortunately be observed on the grain boundary between two differently oriented primary K-feldspar (K1, K2); see Figs. 4a, 4b, 4c, 4d. and 4e. Fig. 4e diagrammatically shows the mineral relationships. The K-feldspar (K1'), replacing the differently oriented K-feldspar (K2), takes the same crystallographic orientation as K1 on which the K1' nucleated, while K2', replacing the differently oriented K1, takes the same crystallographic orientation as K2 on which the K2 is attached. Both replacing K-feldspars, K1' and K2', are distinguished from the primary K-feldspar, K1 and K2, by the presence of residual inclusions of foreign perthite albite (Ab1, Ab2) and vermicular authigenic perthitic albite (Ab1', Ab2').
The K-feldspathization occurred mainly within the space occupied by the primary K-feldspar. Nearly 40 to 50 % of the primary K-feldspar has been replaced by the nibble metasomatic K-feldspar in the rock.
The K-feldspar relationships shown in the Huangnitian quartz syenite can be explained neither as a result of late magmatism, nor as a kind of exsolution from primary K-feldspar (Collins, 1998). The phenomenon should only be interpreted as a result of K-feldspathization, nibbly replacing pre-existing primary K-feldspar in the solid state (Rong, 1982).
The K-feldspathization stops at the contact with quartz and hornblende (Fig. 4b, Fig. 4c; Fig. 5), and is also blocked at the contact with fine-grained plagioclase (Fig. 3b). On that basis, the primary K-feldspar is more apt to be replaced than the other minerals. Fine perthitic albite is moderately replaced, while plagioclase crystals can partly be replaced. During K-feldspar replacement quartz and hornblende are stable and show little to no replacement.
The susceptibility for K-feldspar nibble replacement is as follows.
K-feldspar (easily replaced)
Perthitic albite (moderately replaced)
Plagioclase (partly replaced)
Hornblende, quartz (hardly replaced)
In Huangnitian quartz syenite, along the grain boundaries between two differently oriented K-feldspar crystals (whether original or metasomatic) there are tiny grains of albite, which locally are divided into two rows (Fig. 6a). Each row (0.02-0.05 mm wide) has the same crystallographic orientation as that of the perthitic albite of K-feldspar on the opposite K-feldspar. This phenomenon is clearly noticed under cross-polarized light when a quartz or gypsum plate is inserted. Almost at any contact of plagioclase with disoriented K-feldspar a narrow albite rim, taking the same orientation as the plagioclase, can also be noticed (center left in Fig. 2b), although indistinct (upper right in Fig. 3b), but no albite occurs at the other boundaries (between two plagioclase; between plagioclase or K-feldspar with quartz or hornblende). The albite nibble rim, being of nibble replacement, has widely taken place after the formation of K-feldspathization had ceased, although the nibble albitization here is much weaker than the nibble K-feldspathization.
The fact that remnants of perthitic albite are less abundant than that of perthitic albite in the primary K-feldspar shows that much of perthitic albite has been replaced by K-feldspathization. The average width of albite remnants should be less than the average width of perthitic albite in primary K-feldspar, and that is the case. A few albite remnants in the replacing K-feldspar, however, are wider than the average width of perthitic albite in the primary K-feldspar. Careful observations of the isolated wider albite remnants at high magnification and with a reduced diaphragm show that a Becke line divides the remnant into two parts: an inner nucleus and an outer rim (Fig. 6a). The inner nucleus has a slightly higher refractive index and lower birefringence, indicating a comparatively higher An content (An 6-8?) than that of the outer rim (An ~0), although there is no abrupt boundary between them under normal cross-polarized light. Fig. 6b (diagram of Fig. 6a) shows that the inner nucleus is a metasomatic residue of a perthitic lamella of replaced K-feldspar, whereas the outer albite rim is newly formed around the residue by nibble albite replacement toward the pre-existing (replacing) K-feldspar with different orientation. This phenomenon is consistent with the phenomenon of swapped albite rows. Both are nibble albite replacement.
Huangnitian hornblende quartz syenite has undergone intense nibble K-metasomatism; that is, primary perthitic K-feldspar has been nibbly replaced by metasomatic K-feldspar, followed by weak Na metasomatism in which K-feldspar crystals (both primary and replacing) are nibbly replaced by albite.
Collins (1998) reported an unusual perthitic K-feldspar, containing vermicular plagioclase lamellae, which occurs in monzonite from the Maronia pluton in northern Greece; see Fig. 1 and Fig. 2 in Collins (1998). Because this perthitic K-feldspar is similar to the replacing K-feldspar containing perthitic albite vermicules or stringers in Huangnitian quartz syenite, the author recommends rechecking the thin section of the monzonite to see if there is a non-perthitic K-feldspar with the same crystallographic orientation attached to the perthitic K-feldspar which would act as a nucleation base, and whether remnants of perthitic albite of replaced K-feldspar occur nearby. The remnants of perthitic albite, however, may not necessarily be found in it because the replaced K-feldspar is non-perthitic. Nevertheless, the K-feldspar on which the replacing K-feldspar nucleated might still be noticed. It is undoubted that the "unusual" perthitic K-feldspar from the Maronia pluton is metasomatic in origin.
The presence of vermicular perthitic albite is a characteristic sign for recognizing K-feldspar that has been formed by replacement processes. Swapped rows of metasomatic K-feldspar with vermicular perthitic albite have also been observed on grain boundaries between primary K-feldspars in Hengling pyroxene monzonite (Fig. 7a and Fig. 7b). However, such swapped perthitic albite vermicules are not found associated with those nibbly replacing K-feldspar, which are formed later than albitization. In some older granites, the nibble K-feldspathization, occurring after myrmekitization and albitization (being of nibble replacement, too), is moderately developed, and the vermicular albite stringers in replacing K-feldspar are entirely absent (Fig. 8).
The phenomenon that several small plagioclase crystals are obviously nibbly replaced by different oriented K-feldspar at the border of megacryst K-feldspar or at it inner part gives seemingly the most conclusive evidence that whole megacryst K-feldspar must be formed by K-feldspathization. This is why many researchers are convinced that a K-feldspar megacryst is a porphyroblast rather than a phenocryst because several plagioclase grains have been partly replaced by K-feldspar (Drescher-Kaden, 1948; Collins and Collins, 2002; Roddick, 2002).
This point of view, however, should be reconsidered if the rule of nibble-replacement is admitted that replacing mineral K-feldspar grows on the lattice of a primary K-feldspar at the back, takes the same crystallographic orientation of that primary K-feldspar and grows toward the differently oriented mineral being replaced. Therefore, according to metasomatic intensity observed in thin section, the nibble-replacement of plagioclase by K-feldspar is possibly only locally developed (around plagioclase islands at the grain boundary between plagioclase and K-feldspar), while the main part of megacryst K-feldspar (without relics of plagioclase) is created previously, at least prior to the K-metasomatism, although there is no obvious grain boundary between them due to their identical orientation. This argument cannot easily be verified until a critical phenomenon is noticed that swapped rows of nibble metasomatic albite (Ab1', Ab2'), occurring at the boundary between two differently oriented K-feldspars, is nibbly replaced by K-feldspathization (Fig. 9). Here, the presence of two differently oriented K-feldspar crystals (K1, K2) is obviously a prerequisite condition for the formation of swapped metasomatic albite rows (Ab1', Ab2'). After the nibble albitization had ended, the nibble K-feldspathization began to occur, resulting in the formation of metasomatic K-feldspar (K1", K2") as the scattered albite remnants are in parallel optical continunity with each row of albite nearby.
Furthermore, curiously, notice that numerous tiny albite crystals are scattered around the border between two K-feldspar crystals (Fig. 10a). With a quartz plate inserted, the scattered albite grains are separated into two groups (Fig. 10b). Each group of albite crystals seems to be replaced by K1 and K2 respectively. Therefore, both K-feldspars (K1, K2) look as if they were formed by metasomatism. Nevertheless, because each group of residual albite crystals is in parallel optical continuity with the perthitic albite of the opposite K-feldspar, this phenomenon can only be interpreted by superposition of swapped albite rows being replaced later by the nibble K-feldspathization. The K-feldspars (K1, K2) must be formed before the swapped nibble albitization, while the K-feldspars (K1", K2") surrounding plagioclase islands took place after the formation of the swapped albite rows.
The K-feldspathization occurring in Fig. 10b must be more intensive than that in Fig. 9. The pre-existing primary K-feldspar served as substrate on which the replacing K-feldspar grew while partly replacing the swapped albite and leaving numerous albite relics with their orientations unchanged. Because the nibble metasomatic K-feldspar (K1", K2") has the same crystallographic orientation as that of the primary K-feldspar (K1, K2), no grain boundary exists between them. The nibbly replacing K-feldspar occurring after nibble albitization is characterized by generally less perthitic albite. There is really no obvious difference between the nibble replacing K-feldspar and the primary K-feldspar, if both lack perthitic albite. Primary plagioclase is not apt to be replaced even by intense nibble replacement of K-feldspar, but where it is replaced, a few remnants of plagioclase will likely be preserved in the replacing K-feldspar. Moreover, the grain boundary between two replacing K-feldspars (K1", K2") is not exactly the original one between two primary K-feldspars (K1, K2). The original boundary separating swapped albite rows was destroyed and changed when the two nibbly replacing K-feldspars come into contact after complete replacement of some part of the albite rows. Hence, one can see that the presence of plagioclase remnants in K-feldspar megacrysts does mean that nibble metasomatic K-feldspathization of primary plagioclase has taken place, but only locally around the scattered remnants.
K-feldspar megacrysts may originally contain plagioclase crystals in either zonal or random arrangement (monzonitic texture). The susceptibility for K-feldspar nibble replacement of primary plagioclase is roughly the same throughout the rock. It is unlikely that many plagioclase crystals as well as quartz and biotite would completely be replaced by K-feldspar without any remnants while other plagioclase crystals are only locally replaced and especially, quartz and biotite crystals are unreplaced by K-feldspar everywhere. Therefore, the major part of K-feldspar megacryst (without nibbly replaced plagioclase remnants) being the substrate on which the replacing K-feldspar nucleates, is not metasomatic but magmatic in origin.
The metasomatic K-feldspar nibbly replacing pre-existing K-feldspar is commonly characterized by the presence of vermicular perthitic albite, while the metasomatic K-feldspar nibbly replacing plagioclase is always characterized by the absence of vermicular perthitic albite and less perthitic albite. The former (metasomatic K-feldspar nibbly replacing K-feldspar) occurs prior to albitization, while the latter (metasomatic K-feldspar nibbly replacing plagioclase) takes place later than albitization as far as the author observed.
A series of microscopic evidence, such as the zoned idiomorphic plagioclase with straight Carlsbad- and albite-twinned planes (Fig. 2a), the interstitial quartz without undulated extinction, the straight Carlsbad twin plane of primary K-feldspar (Fig. 5), the swapped rows of tiny albite grains formed along the boundaries between two differently oriented K-feldspars at the end of K-feldspathization, and especially each of the numerous tiny inclusions of perthitic albite (in replacing K-feldspar) with the same consistent (undisturbed) crystallographic orientation, clearly show that the rock remained undeformed. Neither cataclasis nor recrystallization occurred in the rock prior to, during, even after the nibble replacement*.
*Nibble replacement, however, can also occur after the rock has been intensely deformed; for example, myrmekite aggregate may appear at the selvage of relict K-feldspar megacrysts in dynamically deformed, even mylonitized, rocks (Taylor, J. P., Bursnall, J. T., 2003; Cesare, B., 2002). The tiny K-feldspar particles, occurring around a megacryst during strong deformation, serve as nuclei substrate for myrmekitization according to the rule of nibble replacement (Rong, 2002).
It is beyond the imagination for many researchers that introduced potash-bearing gas or fluid can penetrate into such compact undeformed rock and move along so tightly sealed grain boundaries and even permeate through cleavages of rock-forming minerals. Nevertheless, such is a fact that is independent of man's will. Nibble K-feldspar metasomatism has really occurred in such compact undeformed rock. Nibble metasomatism can take place where there is suitable mineral (K-feldspar) prone to be replaced in front of an appropriate mineral (K-feldspar) on which the replacing mineral can nucleate. In addition, the orientation between the two K-feldspar crystals must be different. There is always a tiny crack acting as a passageway for gas or fluid migration and allowing reactions along the grain boundary between two of the identical or similar minerals with different orientations.**
The surface of a grain boundary into which the introduced gas or fluid may infiltrate is the metasomatic active front. As the replacing K-feldspar crystallographically coincides with the back K-feldspar on which the replacing K-feldspar nucleates and grows forward, the metasomatic active front would push gradually and continuously forward towards the opposite replaced K-feldspar, leaving unreplaced perthitic albite as residues in situ. The solid state is always maintained in front of and at the back of the metasomatic active front during metasomatism. The author holds that during nibble K-feldspar metasomatism the unreplaced minerals should never be pushed aside and be parallelly or zonally arranged as Augustithis suggested (Augustithis, 1973).
**There is no crack passageway for gas or fluid migration along the grain boundary between two identical or similar minerals with the same orientation. Therefore, it appears that there is no nibble metasomatism at this spot. On that basis, it may be deduced that perthite and antiperthite cannot be formed through nibble metasomatism of K-feldspar by sodic plagioclase or vice versa. However, the whole perthite (or antiperthite) can be formed by nibble metasomatism as observed in Huangnitian quartz syenite.
Numerous plagioclase crystals may be enclosed either concentrically or randomly in magmatic megacryst K-feldspar. During nibble K-feldspathization, the majority of plagioclase inclusions whose lattices are inclined to the lattice of the host K-feldspar may locally be replaced by K-feldspar, while a few plagioclase inclusions whose lattice has the same orientation as the host K-feldspar would not be replaced. When hydrothermal alteration takes place later, those inclusions with different orientation would be altered and even completely changed into albite, while a few plagioclase inclusions with the same orientation as the host K-feldspar would be preserved undisturbed and even still retain its original zonation. The preservation is not on account of their strong structure (less likely to be fractured) as deduced by Collins and Collins (2002), but just because their parallel lattice boundaries are strictly sealed, preventing introduction of gas or liquid that would allow replacement to occur.
In most places the substrate is an identical or similar mineral, but if such a mineral is absent in the rock, an impurity (lattice defect, interstitial foreign element, etc.) may act as a nucleation substrate. For example, pyrite may nucleate on an impurity and grow in nibble replacement pattern. As evidence of such nibble replacement, pyrite porphyroblasts (1.5 cm in diameter) have been found in cataclastically-deformed granites (Rong, 1982). Inside the porphyroblast, a relict of plagioclase is in optical continuity with a large plagioclase crystal just outside the porphyroblast. Furthermore, the pyrite porphyroblast was euhedrally formed due to its very high crystallizing ability. This is an extreme example that the nibble replacement has taken place when the impurities serve as nucleate substrate in case of lack of same or similar mineral in the rock. Nevertheless, most kinds of nibbly replacing minerals are the same as major rock-forming minerals, and there are sufficient minerals that may act as substrate in the rock. It is unnecessary and unlikely that a replacing mineral will select an impurity as substrate rather than the abundant preexisting identical or similar mineral.
Nibble metasomatism appears at grain boundaries between two minerals one of which is the same mineral as the replacing one or similar to the replacing mineral. The replacing mineral, taking the same crystallographic orientation as the mineral on which it nucleates (just like the case that a mineral crystallizing from a melt or liquid would select a nucleation substrate), gradually grows nibbling at the opposite replaced mineral. The formation of swapped K-feldspar patches, swapped albite rows, and even scattered relics of swapped albite rows in K-feldspar, has clearly revealed the rule of nibble metasomatic growth of rock-forming minerals in undeformed granitoid rocks.
At least four conditions for nibble replacement are required:
a) Metasomatic gas or fluid, introduced from depth;
b) Grain boundaries between two differently oriented minerals
as passageways for moving of gas or fluid;
c) A mineral prone to be dissolved (replaced) in front;
d) A nucleation substrate at the back.
No nibble metasomatism would occur if any one of the above four basic conditions is not met. There is no growing without dissolving. There also is no growing without a nucleation substrate (Rong, 1982). In most places the substrate is an identical or similar mineral, but if such a mineral is absent, an impurity (lattice defect, interstitial foreign element, etc.) may act as a nucleation substrate. For example, pyrite may nucleate on an impurity and grow in nibble replacement pattern. As evidence of such nibble replacement, pyrite porphyroblasts (1.5 cm in diameter) have been found in cataclastically deformed granites (Rong, 1982).
Inside the porphyroblast, a relict of plagioclase is in optical continuity with a large plagioclase crystal just outside the porphyroblast. It is an extreme example that the nibble replacement has taken place when the impurities served as nucleate substrate in case of the lack of a similar mineral in the rock. Because most kinds of nibbly replacing minerals are the same as major rock-forming minerals, there are sufficient minerals that may act as substrate in the rock. It is unnecessary and unlikely that a replacing mineral will select an impurity as substrate rather than the abundant preexisting identical or similar mineralThe unreplaced part of a replaced mineral remains crystallographically unchanged. Scattered remnants of a replaced mineral in a replacing mineral preserve the original crystallographic orientation. A nibbly replacing mineral grows like a silkworm nibbles. The replaced mineral in the replacing mineral keeps the original crystallographic orientation.
The widespread distribution and intensive development of K-feldspathization that occurred in the Huangnitian hornblende quartz syenite can be explained by the comparatively long duration of the metasomatic process and by the recycling of the K, Al, and Si recovered from replaced K-feldspar.
The purpose of this paper is to introduce the phenomenon of nibble metasomatic K-feldspathization, to explain how the mechanism of nibble metasomatism occurs between feldspars, and to present a reasonable explanation for the process and history of metasomatism that occur in granitic rocks. Nibble metasomatic processes in rocks take place not arbitrarily [resulting in the formation of small "replacing" albite crystals chaotically or zonally distributed in K-feldspar or quartz) as deduced by Masgutov (1960) and Beus et al., (1962)], but regularly according to the four above-listed necessary conditions. Therefore, the small albite crystals distributed chaotically or zonally and the host mineral, K-feldspar or quartz, are not metasomatic but magmatic in origin. K-feldspar megacrysts enclosing zoned or even random distribution of perthitic albite lamellae, plagioclase, biotite and quartz inclusions are basically magmatic rather than metasomatic, although local nibble replacement of plagioclase by K-feldspar occurs in the rock.
The real cause and concrete conditions (temperature, pressure, duration, and metasomatic fluid) for the formation of nibble metasomatic K-feldspathization or albitization are still open questions.
I wish to thank L. G. Collins for his honest encouragement and selfless assistance and for script corrections and editorial revisions, offered by him, which have obviously improved the clarity of the article. I wish also to thank Pan Weizhu, Guo Yueheng, Zhou Shuqiang, and Li Yuexiang for providing and making the map of the Huangnitian quartz syenite and its surrounding geology.
For more information contact Jiashu RONG at: mailto:rongjs123@vip.sina.com